An attempt was made to produce animal model of fulminant hepatic failure in the pig by intragastric administration of acetaminophen (N-acetyl-p-aminophenol). The effect of varying doses was observed in phenobarbital-pretreated and untreated animals. Assessment of optimal dose was made in terms of duration of survival and coma in animals exhibiting biochemical and histological evidence of liver necrosis. Significant liver necrosis was observed in enzyme-induced pigs only, whereas in the noninduced pigs respiratory arrest was considered a likely cause of death. Although liver necrosis occurred consistently at a dose of 0.5 to 1.1 g per kg of acetaminophen in phenobarbital-pretreated animals, survival time and coma duration were unpredictable. Only one-fifth of the total number of phenobarbital-pretreated pigs administered acetaminophen in this dose range fulfilled the prescribed criteria for a satisfactory model. There was a good correlation between survival time and duration of coma, although survival time did not correlate with acetaminophen dose, or with any of the following: peak plasma acetaminophen levels, thrombocytopenia, fall in hematocrit or fibrinogen, or alteration in other biochemical parameters. In addition to the unpredictable duration of survival and coma, the occurrence of acute anemia in 60% of the enzyme-induced animals was a major detraction from the possible usefulness of this model to evaluate various therapeutic regimes.